The separation of replication (green) and recombination (red) is lost in cells deficient in the S-phase checkpoint (shown).

Homologous recombination does not rescue stalled replication forks during S phase, report Meister et al. on page 537. Rather they find that replication and recombination are separated in terms of when and where they occur. If inappropriate recombination does occur when replication is stalled, replication forks fall apart.Researchers have debated whether homologous recombination, which is required for double-strand break repair, is also used to resolve stalled replication forks. However, because the S- and G2/M-phase checkpoints share molecular components in budding yeast, the question has been difficult to answer.

Now, Meister et al. test the question in fission yeast, where the checkpoints are molecularly distinct and therefore can be separated. When they added hydroxyurea to wild-type cells, thus depleting DNA precursors and stalling replication, the recombination protein Rad22 did not aggregate on the chromosomes. However, Rad22 foci did appear on chromosomes in cells deficient for the S-phase checkpoint component cds1. Furthermore, transferring the cells to fresh media rescued wild-type cells, but not cds1 mutants. 2D gel analysis of the replication forks showed that the fork structure decayed in cds1 mutant cells, but maintained normal shapes in wild-type cells. In both experiments, cells carrying mutations in the recombination protein rhp51 behaved like wild-type cells.

The researchers conclude that part of the role of the S-phase checkpoint is to prevent inappropriate—and damaging—recombination during the replication process. cds1 is a protein kinase and Meister et al. are now working to learn whether it blocks recombination by directly phosphorylating recombination proteins.